INORGANIC CHEMISTRY.THE Report for 1926 has been prepared in the same manner asprevious Reports and, to save space, the reader’s attention may bedirected to the remarks prefacing the Report for 1925, which areapplicable to this one also.Among subjects which appear to be of special interest are severalcases of apparent transmutation of elements (hydrogen to helium,inercury to gold, lead to mercury and thallium), further work onthe hydrides of boron, the discovery of the missing rare-earthelement (No, 61) ‘. illinium,” and a good deal of work on the effectsof intensive drying, noted in several parts of the Report.Atomic Weights.The technique previously developed for determiningthe density of oxygen has been applied to determine the normaldensity of helium.The value obtained is 0.17846 a t 0’ and760 nim. a t sea level in latitude 45”, g being taken as 980.398.Hence the atomic weight of helium is 4a000 with an experimentalerror affecting the 4th decimal on1y.lA flotation method has been utilised to compare thedensities of samples of boric oxide prepared from six boron mineralsfrom different parts of the world. The boric oxide beads weresealed in glass tubes with the same mixture of dry, inert, organicliquids of known density, and the temperature of flotation of thebeads in each sample was then observed. The mean density offused boric acid a t 18” is 1,7952. The densities of the severalsamples varied from 1.79711 to 1.79404, indicating variations inthe atomic weight of boron from 10.847 t o 10.788.The threesamples for which previous determinations of the ratio BCl, : 3Aghad given the values 10.841, 10.825, and 18.818, gave relative valuesdeduced from the densities 10.847, 10423, and 10.818.2Silicon. A similar flotation method, using glass floats of appro-priate density, previously calibrated by determination of theirflotation temperatures in a mixture of organic liquids of knowndensity, has been applied to compare the densities of, and hence the1 G. P. Baxter and H. W. Starkweather, Proc. Nut. Acod. Sci., 1925, 11,231; 1926,12, 20; A., 1925, ii, 1045; A,, 1926, 233.* H. V. A. Briscoe, P. L. Robinson, and G . E. Stephenson, J., 1926, 70;A., 219; compare Ann. Repvrte, 1926, 22, 43.Heliunz.Boron50 ANNUAL REPORTS ON TEE PROQRESS OF CHXMISTRY.atomic weights of the samples of silicon in, preparations of silicontetrachloride derived from five precisely known localities in Canada,the United States, Sweden, Scotland, and France, and subjectedto a rigorous purification by fractional distillation under exclusionof moisture, first a t atmospheric pressure and afterwards in avacuum.The extreme values for the atomic weight of silicondeduced from the densities of silicon tetrachloride are 28.058 and28.063, the maximum variation thus being one part in 6000 parts,and the probable error of individual relative values of the atomicweight considerably less than this. These results are held to showthat any variation in the atomic weight of silicon from differentsources is substantially less than one unit in the second decimalplace,3 and confirm and extend the conclusion deduced from pyknom-etric measurements of the density of tetraethylsilicane by Jaegerand Dijl~stra.~Chlorine.Determinations are recorded of the ratio AgCl : Ag,using samples of chlorine derived from the sea, from three mineralsof non-marine origin, viz., apatite, wernerite, and socialite, and froma meteorite. The mean value of the atomic weight for chlorine incommon salt and the three terrestrial minerals is 35.457 & 0.0002,whilst that of the meteoritic chlorine is 35.458 5 0-0005. Thesedeterminations afford further evidence, much more precise thanthat hitherto available, that the atomic weight of chlorine doesnot vary with its source.6Metallic germanium derived from the germaniumtetrachloride previously used for atomic-weight determinations,Gwas converted into germanium tetrabromide, and this compound,after purification by 13 fractional distillations in a vacuum, wasused for determinations of the ratios GeBr, : 4Ag, GeBr, : 4AgBr.The mean of 32 analyses gave Ge = 72.60, a value identical withthat given by analysis of the tetrachloride.'Silver. A thorough investigation has shown that silver oxideis much more stable than has hitherto been supposed, and that,when it is prepared by precipitating silver nitrate with baryta underrigid exclusion of organic matter and carbon dioxide, it may be3 P.L. Robinson and H. C. Smith, Nature, 1926, 118, 303; A., 999; J.,1926, 1262; A., 771; H.V. A. Briscoe and P. L. Robinson, Nature, 1926,117, 377; A,, 331.4 F. M. Jaeger, 2. Elektrochem., 1926, 32, 328; A., 870; compare Ann.Reports, 1925, 22, 44.6 W. D. Harkins and S. B. Stone, J. Amer. Chem. SOC., 1926, 48, 938;A., 553.6 Ann. Reports, 1924, 21, 28.7 G. P. Baxter and W. C . Cooper, J. Phy&cal Chem., 1925, 29, 1364; A.,Cermaniurn.1926, 5INORGANIC CHEMISTRY. 51heated in a current of pure air at 120' for a week without decom-position, and thereafter will yield a white chloride on treatmentwith hydrochloric acid. By heating pure silver oxide thus pre-pared in a silica tube a t 350---100', in a current of pure dry air, 6direct determinations were made of the ratio Ag,O : Ag, giving amean value for the atomic weight of silver Ag = 107.864 f 0.0013.The original paper must be consulted for interesting details of thepreliminary investigation into the stability of silver oxide and of theapparatus and methods used in the actual determinations.8 Ninedeterminations of the silver remaining after ignition of silvercarbonate, prepared and dried under the conditions found to giveB minimum of decomposition, gave a mean value for the atomicweight of silver Ag = 107.86.*Lead.I n the course of unsuccessful attempts to obtain someseparation of the isotopes present in ordinary lead by irreversiblevolatilisation and by the Grignard process, 18 values for the ratioPbCl, :2Ag were obtained, giving a mean value for the atomicweight of lead Pb = 207.217 f <O.OO1.loFour determinations of the ratio PbC1, : 2Ag upon lead extractedfrom a specimen of uraninite from the Black Hills, South Dakota,gave a mean value for the atomic weight of lead Pb = 206.07.When a correction is applied for the known thorium content of themineral, it appears that the atomic weight of uranium-lead in thisspecimen is Pb = 206-02. It is of interest that the high lead-uranium ratio, 0.23, of this relatively pure uranium-lead indicatesan age for the mineral of a t least 1500 million years.llTitanium.I n continuation of work previously reported, 17determinations of the ratio TiCl, : 4Ag, made on material from thelater stages of the fractionation of the tetrachloride, yielded resultsfor the atomic weight of titanium lying within the limits 47483-47.922 and giving the mean value Ti = 47.90.11a.GTOUP 0.A development in the methods of detecting helium hss affordedevidence of the production of that gas from hydrogen.By remov-ing the relatively condensable gases with charcoal and liquid air,burning hydrogen with excess of oxygen on a platinum or palladiumcatalyst, absorbing the residual oxygen with charcoal, and examining* H. L. Riley and H. B. Baker, J . , 1926, 2510; A., 1190.9 G. H. Jeffrey and A. W. Warrington, Chem. News, 1926, 132, 373; A.,T. W. Richards, H. S. King, and L. P. HaI1, J . Amer. Chem. Soc., 1926,T. W. Richards and L. P. HalI, i b d , p. 704; A,, 449.11s G. P. Baxter and A. Q. ButIer, ibid., p. 3117; compare Ann. Reports,694.qS, 1530; A,, 771.1923, 20, 3062 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the residual gas spectroscopically in a minute glass capillary, itis possible to detect quantities of helium as small as 10-8 or 10-9 C.C.and thus to detect the gas evolved from active thorium precipitates,and determine the quantity in quite small samples of natural gas.When hydrogen free from helium is passed over heated palladium,the issuing gas is found to contain helium, and the quantity of heliumappears to be increased when the hydrogen is allowed to remainin contact with palladium-black, spongy palladium, or palladisedasbestos a t the ordinary temperature.Although the behaviour ofthe catalyst is somewhat irregular, and its activity diminishes withtime, there is a rough proportionality between the duration ofcontact of the hydrogen with the metal and the quantity of heliumfound thereafter.Catalysts inactive to hydrogen, and occasionallycatalysts which readily absorb hydrogen, do not yield helium, but,in general, a catalyst active towards hydrogen does produce helium,A catalyst which has become inert may be revived in the usualmanner, and may then produce helium. Specimens of finely.divided palladium which have been preserved for a time a t theordinary temperature always yield helium on heating. There is,of course, some neon in this gas, but the ratio of helium to neon isusually much greater than that in air. These observations all leadto the conclusion that the palladium in this case acts purely as acatalyst for the conversion of hydrogen into helium, and this viewis confirmed by the further observation that a similar, although less,effect is produced by platinum.12Measurements of the surface tension of liquid helium in contactwith its saturated vapour by the method of capillary rise have shownthat the molecular surface tension increases linearly with fall oftemperature down to 2.4" Abs., and thereafter approaches a constantvalue in the neighbourhood of 1.5" Absi.13 By observing the absolutetemperatures and pressures a t which a tube system, containingliquid helium, became blocked, it has been inferred that the fusioncurve of helium is described by the following points : 1.l0, 2 6 atm.;2.2") 50 atm, ; 3-2") 86 atm. ; 4.2") 150 atm. When helium was frozenin a glass tube, no boundary surface was visible between the solidand liquid phases, whence i t appears that the refractive indices ofsolid and liquid helium must be closely ~imi1ar.l~ It has been shownthat quartz glass is permeable to helium under a pressure of 100 atm.a t laboratory temperatures, whereas under the same experimentalconditions no permeability to hydrogen could be detected.1512 F.Paneth and K. Peters, Ber., 1926, 59, [B], 2039; A., 1077.13 A. T. van Urk, W. H. Keesom, and H. K. Onnes, Proc. K. Akad. Wetenach.l4 W. H . Keesom, Compt. rend., 1926, 188, 26, 189; A., 892, 893.H. M. Efsey, J . Amer. Chem. Soc., 1926, 48, 1600; A., 895.Amsterdam, 1925, 28, 958; A,, 1926, 568INORGANIC CHEMISTRY. 53Croup I.Some interesting work is reported on the reduction of aqueoussolutions of metallic salts by hydrogen under pressure.Withsolutions containing 3-30% of platinum chloride, the yield ofplatinum in unit time increases with temperature and with pressureof hydrogen, whilst the proportion of the total platinum precipit-ated increases with diminishing initial concentration of the solution.Presence of iron and nickel salts and mineral acids greatly retardsor inhibits the reduction.16 The action of compressed hydrogenon hot copper sulphate solution yields, first the basic saltCuSOp,2Cu( OH),, then cuprous oxide, and ultimately copper, thequantity of which increases with the amount of free sulphuric acidpresent. At 150°, there is some reduction of sulphuric acid andthis facilitates the separation of basic salts and cuprous oxide;a t higher temperatures copper sulphide is produced.This reduc-tion of sulphuric acid is accelerated by the precipitated copper.Chromic acid, alone or in the presence of sulphuric acid, is reducedt o the oxide, Cr,O,,H,O ; whilst potassium dichromate, acidifiedwith sulphuric acid, a t 300" and 80 atm. of hydrogen, yields smallviolet-grey crystals of a salt, K20,2Cr,0,,3SO,,H,O, insoluble in acidor alkali. From nickel formate, under relatively drastic conditions,anhydrous, crystalline nickelous oxide is produced ; lower temper-atures and pressures give a quantitative yield of metallic nickel.Phosphoric acid is not reduced a t 350°, but lead hydrogen orthophos-phate is reduced to lead hydrogen phosphite, hypophosphorous acid,and colloidal lead oxide.Red phosphorus a t 200" and 90 atm. isconverted into black phosphorus, but under milder conditions ityields phosphine and phosphoric acid. Many other interestingreactions are described in the original papers.17A good deal of evidence converges on the view that many of theso-called metallic hydrides are in fact not stoicheiometric compoundsbut solid solutions approximating closely thereto. Even calciumhydride, usually regarded as a typical salt-like hydride of the typeformed by the alkali and alkaline-earth metals, is found to containless hydrogen than is required by the formula CaH,, and t o show,even a t 20", a measurable hydrogen pressure, steadily increasingwith time, even after 9 days. On removing hydrogen from calciumhydride, the substance slowly separates into two portions, oneapproximating closely to CaH,, the other poorer in hydrogen. Simi-larly, it is found that dry copper hydride always contains lesshydrogen than is required by the formula CuH and loses hydrogen16 V.Ipatiev and A. hdreevBki, C m p t . rend., 1926, 183, 51; A., 921.l 7 V. Ipatiev and others, Ber., 1926, 69, [B], 1412; A,, 921; V. N. Ipatievend B. A. Mouromtsev, Compt. rend., 1926, 183, 505; A., 111484 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.progressively when heated.18 Observations on the absorption ofhydrogen by praseodymium, neodymium, zirconium, and thorium,and on the dissociation of the products, indicate that these hydridesalso, as well as those of cerium and lanthanum, are solid solutionswhich, under favourable conditions, may approximate to, but neveractually attain, the composition and stability of stoicheiometriccompounds.lgA thorough investigation of the action of copper on concentratedsulphuric acid has shown that whilst a t all temperatures from 16"to 270" the completed reaction is represented by the equation Cu+2H,S04 --+ SO, + CuSO, + 2H,O, there actually occur at leastfour different reactions :(i) 5cu + 4H,SO, = cu,S + 3CuS0, + 4H,O ;(ii) Cu,S + 2H,SO, = CuS + CuSO, + 2H,O + SO, ;(iii) CuS + 2H,S04 = CuSO, + 2H,O + SO, + S ;(iv) S + 2H,SO, = 2H20 + 350,.In the temperature range 100-120°, action (i) is much more rapidthan actions (ii) and (iii) ; hence sulphide formation is particularlyevident.At 270", actions (ii) and (iii) are so rapid that the presenceof sulphides as intermediate compounds is not readily detected.The crystalline deposit formed is anhydrous copper sulphate, whichis white when formed a t high temperatures, but a t lower temper-atures is grey, owing to inclusion of black sulphides; micro-scopic examination of the crystals gave evidence that this salt isdimorphous.20Evidence is adduced that discrepancies in the literature relativeto basic copper sulphates are attributable to the formation of highlystable intermediate compounds, and that the basic salt obtained byboiling solutions of copper sulphate for a short time is5CuS04,9Cu( OH),,2H20.This salt can easily be produced in quantity if the acid producedon hydrolysis is removed as it is formed by interaction with sodiumnitrite present in the solution.On prolonged boiling with water,this basic salt or copper sulphate yields CuS04,2Cu(OH),. Byhydrolysis of copper sulphate in solutions over a small range of con-centration near saturation, a new basic salt, 2CuS04,Cu( OH),,4H20,was obtained which has apparently eluded previous observers owing toits decomposition by water. Two other basic salts, CuS04,3Cu( OH),and 2CuS04,3Cu( OH),, were recognised as definite compounds.1* a. F. Huttig, 2. angew. Chem., 1926, 39, 6 7 : A., 354; G. F. Huttig andF. Brodkorb, 2. anorg. Chem., 1926, 153, 235, 309; A., 694, 809; see alsoH. Miiller and A. J. Bradley, J ., 1926, 1670.A. Sieverts and E. Roell, ibid., 150. 261; 153, 289; A., 356, 810.*O C. W. Rogers, J., 1926, 264MORGANIU OHEMISTRY. 55The original paper contains a very thorough discussion of the wholeliterature of the basic copper sulphates in the light of the presentwork and merits careful study.21Reinvestigation of the cuprous alkali thiosulphates has disclosedthe existence of the ammonia compounds Cu,S20,,2K2S203,NH3 andCu,S203,Na2S,03,2NH,.22Silver perchlorate is unique among typical metallic salts in beingreadily soluble in toluene, the solution saturated a t 25' containing50.3% of the salt : below 22.6" the solid phase in equilibrium with thesolution is AgClO,,C,H,, and the solubility falls off rapidly a t lowertemperatures. These observations are incidental to an examinationof the ternary system-silver perchlorate-toluene-water-for theresults of which the original paper must be consulted.23Further application of the Steele-Grant microbalance and themethods previously described has shown that optimum concen-trations of chlorine exist for the chlorination of both fresh andpreviously-chlorinated silver films, that photochemical decom-position of silver chloride and silver iodide a m s may proceed to theextent of 94-95y0, and that there is no evidence of the formation ofsubchloride or ~ubiodide.~~Group I I .Compact masses of beryllium have been prepared by electrolysingthe double fluoride a t 1200" in a graphite pot with graphite anodespreviously impregnated with the salt.The rotating cathode has aberyllium tip held in a water-cooled holder and is slowly raised aselectrolysis proceeds: thus a solid rod of beryllium is obtained.The crude metal is free from impurities except about 0.05% each ofiron, carbon, aluminium, and magnesium and about 0.005~0 nitrogen;by sublimation in beryllia pots, carbide-free metal containing lessthan 0.02% iron was obtained. Beryllium has d 1-84 and m. p,about 1280" ; it takes a high polish, resists atmospheric corrosionwell, does not readily ignite, is non-ductile, and has a Brine11 hard-ness usually about 140 but in one annealed sublimate as low as 90.,lInvestigations of the viscosities of solutions of beryllium sulphate,selenate, and oxalate containing dissolved beryllia, and of theconductivities of neutral and basic solutions of beryllium chlorideCompare, however, H.T. S. Britton, ibid.,p. 2868.A. Benrath, H. Niehaus, H. Meekenstock, and H. Essers, 2. unwg.Chew., 1926, 151, 31; A., 367.A. E. Hill and F. W. Miller, J. Amr. Chern. Soc., 1925, 47, 2702; A.,1926, 26.E. J. Hastung, J., 1925,127, 2691; 1926, 1349; compare Ann. Repwta,1922, 19, 43; 1924, 21, 35.s1 A. C. V i v h , Trana. FurMEay Soc., 1926, 22, 211; A,, 1114.a1 G. Fowles, J., 1926, 184556 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.and oxalate, yield results consistent with the view that the sulphateand selenate solutions contain a complex cation, Be,xBeO, where xon the whole is less than 4. In the course of this work, berylliumbensenesulphnate and p-bluenesulphonate were obtained as crystal-line salts with 4H,0.32By using a crucible, stirrer, and thermometer-sheath made of a5 : 2 mixture of bole and alumina, which softens only a t 1560" andis not attacked by the melt, it has been possible to make a thermalanalysis of the calcium and magnesium silicides.Two calciumsilicides, CaSi, m. p. 1220", and CaSi,, m. p. 1020°, are formed,and there is some evidence for a third compound, Ca,Si, m. p. 920".Both CaSi and CaSi, are decomposed by water, yielding spontane-ously inflammable silicon hydrides. Thermal analysis of mag-nesium-silicon melts shows only the compound Mg,Si, m. p. 1070",already known, but when the melt is rapidly cooled from above1050", a second silicide, MgSi, is obtained, which is also formed byvolatilisation of magnesium when Mg,Si is kept at temperaturesabove 600".Above 1100", both silicides dissociate, yielding theelements .aA thermometric study of the setting of plaster of Paris, in whichthe maximum on the time-temperature curve was taken as thetime of setting, shows that setting occurs in two stages : (a) aslightly exothermic absorption with contraction in volume, and ( 6 ) amarkedly exothermic reaction between the absorbed liquid and theabsorbent with an attendant increase in volume. The acceleratingeffect of cations in this change is in the order K>NH,>Na>Li;Zn = Cu>Mg : the decelerating effect of anions is in the orderI > NO,> Br >Calcium sulphate, on being heated with carbon at 900" 0s withhydrogen a t 600400", yields calcium sulphide, but above 900" thisproduct reacts with undecomposed sulphate, yielding lime andsulphur dioxide ; a t higher temperatures, interaction occursbetween carbon monoxide and sulphur dioxide, producing sulphur,with carbon oxysulphide as a by-product.35 Simple dissociation ofcalcium sulphate begins at 960" and the dissociation pressure reaches97 mm.a t 1230"; in an equimolecular mixture with amorphoussilica, dissociation begins at 870" and produces a pressure of 817 mm.8% N. V. Sidgwick and N. B. Lewis, J . , 1926, 1287; see general discussionof co-ordinated additive compounds of beryllium, R. Fricke, 2. angew. Cltem.,1926, 39, 317; R . Fricke and 0. Rode, 2. anorg.Chem., 1926, 152, 347;R. Fricke and L. Havestadt, ibid., p. 357; A., 368, 694, 695.33 L. Wohler and 0. Schliephake, ibid., 1926, 151, 1; A., 368.a4 H. A. Neville, J . Physical Chem., 1926, 30, 1037; A,, 899.36 J. Zawadzki, J. Konarzewski, W. J. Lichtenstein, S. Szymankiewicz,and J. Wachsztejnski, Rocz. Chem., 1926, 6, 120, 236; A., 923INORGANIC CHEMISTRY. 57a t 1280". The effect of alumina and ferric oxide has also beenstudied.3sWhen baryta is heated with cupric sulphide or lead sulphide itbrings about a partial reduction to metal (72;/, in 1 hr. a t 1150"in the case of copper) according to the equations of the type : 37CuS + BaO --+ CuO + BaS ; 4CuO + Bas -+ BaSO, + 4Cu.From a study of the solubility curves and the composition of thesolid phases in the system M,O,-BaO-H,O a t 20°, only two bariumaluminates could be isolated, wiz., 2BaO,A1,O3,5H,O, stable insolutions containing from 3.5 to 2.10,', of barium oxide, andBa0,A1,03,6H,0, stable in concentrations of from 2.1 to 1.2% ofbarium oxide. Sevcral coinpounds described in the literature werenot obtained.The former compound is rapidly decomposed by waterinto the latter, which is also decomposed by a large excess of waterinto barium hydroxide and gelatinous aluminium hydroxide.38The cooling curves of calcium amalgams and their microstructureswhen frozen on glass surfaces are consistent with the existence ofthree compounds, CaHg,, CaHg5, and CaHg,,, and the second ofthese may be obtained in relatively large crystals by pouring theamalgams into water.39Stock has directed attention to the possibility that workers exposedto the vapour given 08 by mercury a t laboratory temperatures maycontract very serious mercury poisoning, unless there is an ex-tremely good system of ventilation.The condition of chronicpoisoning thus developed can only be cured by several years'abstention from all work involving the use of mercury. Someworkers deny the possibility of such poisoning, but others can con-firm Stock's experience>O It seems possible that much depends onpersonal idiosyncrasy, so that, whilst some persons are relativelyimmune, others may in fact contract chronic mercurial poisoningunder ordinary laboratory conditions ; if so, a clear case exists forthe exercise of greater precaution than has been customary in theuse of mercury.a6 (Mlle.) G.Marchal, J. Chim. phys., 1926, 23, 38; A., 359; BUZZ. SOC.chim., 1926, [iv], 39, 401; A,, 487.37 W. Biltz and E. von Muhlendahl, 2. anorg. Chem., 1925, 150, 1 ; A.,1926, 136; I. A. Hedvall, Svensk Kem. Tidskr., 1923, 37, 166; from Chem.Zeatr., 1925, 11, 1946; J. A. Hedvall and E. Norstrom, 2. anorg. Chem.,1926, 154, 1 ; A., 368, 695.sB G. Malquori, Uazzetta, 1926, 56, 5 1 ; A,, 810.39 A. Eilert, 2. anorg. Chem., 1926, 161, 96; A,, 356.40 A. Stock, 2. angezo. Chem., 1926, 39, 461; A. Schmidt, ibid., p. 786;G . Pinkus, ibid., p. 7 8 7 ; H. Reihlen, tbid., p. 788; F. Gradenwitz, zbid., p.788; L. Wolff,ibid., p. 789; A. Stock, ibid., p. 790; K. Hofer, ibid., p.1123;A,, 707, 815, 122358 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.I n heating mercury in a sealed silica bomb, a marked distinctionbetween the liquid and vapour phases was observed immediatelybefore the bomb burst above 1000" ; hence the critical temperatureof mercury certainly lies above that temperature.41By using the metal as the liquid in a V-shaped silica manometertube, closed a t one end by a sealed-in thermocouple sheath andopen a t the other end t o a measured applied pressure of nitrogen,measurements have been made of the vapour pressures of mercuryfrom 200" (14 mm.) to 397.5" (1490 mm.), of cadmium from 500"(14 mm.) to 836" (1536 mm.), and of zinc from 625" (16 mm.) to982' (1517 mm.).43Group I I I .Some interesting additions have been made to our knowledge ofthe boron hydrides. Pure diborane has d-112' 0.447 (liquid), d-183'0.577 (solid), and slowly decomposes a t the ordinary temperature,yielding pentaboron hydride, B5Hll, b. p.0°/57 mm., m. p. about- 129". Unlike the higher hydrides, diborane is not oxidised by airor oxygen a t 15": it is hydrolysed by excess of water to boricacid and hydrogen. Ammonia with B5H1, yields hydrogen and scompound, B,H,(NH,),, which closely resembles that formed fromB5H,.43 This ammine when heated gives a compound, B,N,H,.Diborane and ammonia a t 15" give an additive compound,B,H6,(NH3),, which in solution behaves as an alkaline solution ofdiborane, when,heated in a sealed tube gives B,N3H,, and reactswith hydrogen chloride according to the scheme : B,H6(NH3), +2HC1 = B,H,Cl,(NH,), + 2H,. The action of ammonia ondiborane a t about 200" leads to the replacement of the hydrogenatoms by amino- or imino-groups, SO that when excess of ammonia,is used, the ultimate product is boroimide, B,(PU'H),.If theammine, B,H,(NH,),, is similarly heated, an analogous actionoccurs, but the amount of ammonia available is insufficient forcomplete replacement of the hydrogen atoms, and the main and-onlyvolatile product is the compound B3N3H,, mixed with non-volatilecondensed substances of composition between (BNH,), and (BNH),.The compound B3N,H6 has b. p. 0"/84.8 mm., m. p. -58.0", d-651.00 (solid), d-57' 0.898 (liquid), do' 0.824 (liquid), and is unusuallystable. At a high temperature, it decomposes into the compound(BNH), and hydrogen.It is indifferent to oxygen. In cold water,it dissolves to an initially neutral solution, which gradually becomesalkaline ; warm water causes quantitative hydrolysis t o boric acid,4 1 L. A. Sayce and H. V. A. Briscoe, J., 1926, 957.4 1 C. H. M. Jenkins, Proo. Roy. Soc., 1926, [A], 110, 456; A., 333.'3 Ann. Report.?, 1924, 21, 37; compare {bid., 1923, 20, 38INORGANIC CHEMISTRY. 59ammonia, and hydrogen. Ice-cold water yields the hydmte,B,N3H6(H20),, which is converted by anhydrous hydrogen chlorideinto the compound, B3N3H,Cl,(H,0),, and hydrogen. The sub-stance B,N3H, and hydrogen chloride slowly yield the non-volatilecompound, B3N3H6(HC1),. Ammonia is absorbed by the compoundB,N,H,, but the reaction appears complex.The behaviour of theNHvBH compound is best expressed by the constitution BH<NH,BH>NH.The action of iodine on diborane affords mainly boron tri-iodide,m. p. 48-1", and oily products. Diborane is readily converted byhydrogen iodide in the absence of a catalyst a t 50" into the unstableiodo-derivative, B,H,I, m. p. -110", b. p. 0"/78 mm., 2.0(solid), d-loS. 1.8 (liquid). Even a t low temperatures it decom-poses moderately rapidly into diborane and boron tri-iodide. It israpidly and quantitatively hydrolysed by water to boric acid,hydrogen iodide, and hydrogen. It is converted by sodium amal-gam a t -35" into the hydride, E4H10, for which the constitutionBH,-BH,*BH,*BH, is rendered probable if the '' ethane " structurefor diborane be accepted.A study of the infra-red absorptionspectrum of diborane and its X-ray analysis affords strong evidencefor the constitution BH,*BH,, and the following structures areinferred €or the other hydrides : B,H, = BH,.[BH],*BH, ; B5H,1=BH,-[BH],*BH,.A modified nomenclature for the boron hydrides is suggested,according to which the term borane is restricted to the "limithydrides " containing tervalent boron, e . g., B2H, = diborane,B3H5 = triborane, B,,H,, = decaborane. The hydrides richer inhydrogen (the only hydrides already isolated) are termed '' hydro-boranes," thus, B,H,, dihydrodiborane ; B,H,, dihydropentaborane ;B5Hl1, tetrahydropentaborane.&Boron sulphide is prepared by heating boric oxide with aluminiumsulphide in a current of nitrogen a t 1200-1300" : silicon disulphide(m.p. 1090", d 2.02) is similarly obtained when sand is substitutedfor boric oxide and is separable from accompanying silicon mono-sulphide by reason of its different ~ o l a t i l i t y . ~ ~The freezing-point diagram of the system boron trifluoride-hydrogen sulphide shows two eutectics a t -148" and -140", with22% and 5374, of hydrogen sulphide, respectively, between whichlies a maximum a t -137" indicating the existence of a compoundBF,,H,S which is apparently much dissociated a t its melting point44 A. Stock and E. Pohland, Bw., 1926, 59, [B], 2210, 2215, 2223; A.46 E. Tiede and M. Thimann, ibid., p. 1703; A,, 1112.BH3*BH2.BH*BH,*BH3 ; BGH1, = BH3fBHI4*BH3 ; B1,HId =Stock, ibid., p.2226; A., 1217, 121860 ANNUAL REPORTS ON THE PROQRESS OF CHEMISTRY.and should therefore not hinder the separation of the constituentsby fractional di~tillation.~~By rapid manipulation and exact adherence to specified con-ditions, aluminium hydroxide is obtained in three forms, c(, p, andy, which behave as distinct chemidal compounds, all of the formulaAl(OH),. By heating any one of these in a sealed tube at 250"with 10% ammonia, aluminium metahydroxide, AlO.OH, isobtained, having neither basic nor acidic properties but showingremarkable ability to adsorb enzymes ~electively.~'Several fluoroaluminates have been prepared and all these can berepresented, in conformity with other known aluminium compounds,with a.co-ordination number of 6, if in certain cases the nucleusis doubled, e . g., [F,Al<~>AlF4](N2H4),. A similar mode offormulation is applicable to a new potassium aluminium fluoride,A1F,,2KF,2H2O, and also t o a number of complex organic fluoridesof iron and chromium.48Thallium metasilicate, Tl,SiO,, is obtained as a white, amorphousprecipitate when a 4% solution of sodium metasilicate is slowlyadded to a solution containing 20,/; of thallous nitrate and 4% ofthallous hydroxide. An excess of thallous hydroxide must alwaysbe present, as the silicate readily undergoes hydrolysis. Thalliumorthosilicate, Tl4SiO4, is prepared (i) by adding a concentratedsolution of sodium metasilicate to a boiling 16.5% solution ofthallous hydroxide, a crystalline, canary-yellow precipitate beingobtained] which consists of the orthosilicate together with a littlemetasilicate ; (ii) by boiling thallium metasilicate with an excess of0-75N-thallous hydroxide solution ; (iii) by shaking finely-divided,precipitated silica with an excess of thallous hydroxide solution.Both the above thallium silicates are anhydrous] in contradistinctionto the metasilicates of sodium and lithium, which contain 9 mols.and 1 mol.of water, respecti~ely.~~Cerium, lanthanum, praseodymium, neodymium, and samariumhave been prepared as chemically pure metals by electrolysis ofthe fused chlorides in graphite cells, using graphite anodes.5046 A. F. 0. Germann and H. S. Booth, J. Physical Chem., 1926, 30, 369;A., 475.4 7 R.Willstatter, H. Kraut, and 0. Erbacher, Bet., 1025, 58, [BJ, 2448,2458; A., 1926, 34, 35.4 * H. Weinland, I. Lang, and H. Filrentscher, 2. anorg. Chem., 1925, 150,47; A,, 1936, 136.49 K. A. Vesterberg and C . U. Willers, Arhv Kemi, Mzn., Geol., 1926, 9,No. 26, 1; A., 695.E. E. Schumacher and J. E. Harris, J. Amer. Chem. Soc., 1926, 48,3108; see also, for the similar preparation of yttrium, A. P. 'Thompson,1%'. B. Holton, and H. C . Kremers, Tram. Amer. Electrochem., Soc., 1926, 49,161; A., 489INORGANIC CHEMISTRY. 61Further work has confirmed the individuality of the black oxide ofpraseodymium, Pr,O,,, which has d20' 6-61, and is not dissociated upt o 900" : it is regarded as a salt-like compound of R,O, with a higher0xide.~1Persistence in the search for element 61 has been rewarded bysuccess.Examination of the L-series X-ray lines of carefully puri-fied samples of rare earths showed a single faint line in the correctposition for L, 61.52 Fractional crystallisation of the cerium earthsas the magnesium double nitrates concentrates element 61 betweenneodymium and samarium, both having broad absorption bandscapable of masking any bands due to element 61, and fails to give asufficient concentration of this element for certain detection by theX-ray spectrum. If, however, these earths be fractionated as thebromates, element 61 is separated from neodymium and samariumby terbium and gadolinium, respectively. Terbium has but oneabsorption band and gadolinium has none ; hence it became possibleto observe faint bands a t 6700 and 5905 A.and stronger bands a t5830,5816, and 4520 if. attributed to the new element. The X-rayemission spectra of the samples showing these bands gave linescorresponding closely with the calculated positions for La,, and LB,,of element 61 ; the authors therefore claim to have discovered thiselement and propose for it the name " illinium." 53Group I V .Pure carbon tetrafluoride has been isolated by liquefying andfractionating the gases evolved a t a carbon anode used in theelectrolysis of fused beryllium fluoride; it differs from the com-pounds previously described as carbon tetrafluoride, which wereprobably mixtures. It has b. p. -150°, do' 3.034 (air = l),M 87-4, is decomposed by sodium at 500", by calcium at 600°, andundergoes partial dissociation a t 1100'.54Further work on the melting point of graphite, with an improvedapparatus in which higher pressures of argon and a closer approxim.ation to black-body )' radiation could be obtained, has given them. p. 3845" & 45' Abs., and shorn that there is no systematicvariation of m. p. over the pressure range 2-9 atm.655 1 W. Prandtl and K. Huttner, 2. anopg. Chem., 1926, 149, 236; A., 1926,137; compare Ann. Reports, 1924, 21, 40.52 C. J. Lapp, R. A. Rogers, and B. S. Hopkins, Physical Rev., 1925, [ii],25, 106; A., 1926, 1083; compare Ann. R e p r f s , 1924, 21, 39.53 J. A. Harris and B. S. Hopkins, J . Amer. Chem. SOC., 1926, 48, 1585;A., 810; J. A. Harris, L.F. Yntema, and B. S. Hopkins, aid., p. 1594; A.,780; compare L. Rolla and L. Fcrnandes, Qazzetta, 1926,58, 435; A,, 1083;13. Brauner, Nature, 1926, 118, 84; A., 780.54 P. Lebeau and A, Damiens, Compt. rend., 1926, 182, 1340; A., 710.6 6 E. Ryschkewitsch and F. Merck, 2. Elelctrochem., 1926, 32, 42; A,,232; compare H. Herbst, Phy9ikaZ. Z., 1926, 27, 366; A., 67062 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.In the presence of finely-divided palladium and absence of air,carbon monoxide is oxidised by water to carbon dioxide, withsimultaneous production of palladium hydride. Owing to thoisolation, in previous work, of small amounts of barium formate,it has been assumed that the primary action consisted in the additionof water to carbon monoxide, with formation of formic acid, which issubsequently decomposed into carbon dioxide and hydrogen.Theexperimental conditions have therefore been arranged in such amanner as to remove formic acid, if produced by hydration of carbonmonoxide, from the dehydrogenating action of palladium, using forthis reason alkaline solutions containing a sufficient amount of ethylalcohol; the presence of alcohol or sodium hydroxide has noinfluence on the ‘‘ oxygen-free combustion,’’ of carbon monoxide.Under these conditions, carbonate is immediately formed in thesolution, and therefore must be produced directly from carbonmonoxide; its production is accompanied by an increase in thehydrogen content of the palladium. Sodium formate, added beforethe experiment, is found unchanged in amount a t its conclusion.The production of carbonate is invariably less, and that of hydrogengreater, than corresponds with the volume of carbon monoxideabsorbed.This is caused by the very slow action of palladium onethyl alcohol, from which hydrogen is withdrawn, with productionof acetaldehyde, which, however, never Ieads to that of recognisableamounts of carbon dioxide.56The energy of the spark discharge necessary to ignite mixtures ofcarbon monoxide (2 vols.) and oxygen (1 vol.) varies with themoisture content of the gases; for gases saturated at 17” andcontaining 2% of water the energy is 4.6 x 10-3 joule; for gasessaturated at O”, containing 0.6% of water vapour (by vol.) it is29.0 x lo3 joule ; when the gases are dried over calcium chlorideand contain 0.03% of water, it is 126 x lo9 joule; whilst after dry-ing the gases for 6 months over phosphorus pentoxide, the minimumenergy required for ignition is about 0.3 joule.When the gases aredry, the explosion is softer and at atmospheric pressure the reactionis incomplete, its extent depending partly upon the energy of thedischarge. At higher pressures, the gases are more readily ignitedand the action proceeds further, so that above 10 atm. ignition isinstantaneous and the reaction is complete. Spectrograms ofexplosion flames in dry mixtures at 25 atm. show a complete absenceof the steam lines characteristic of the ordinary combustion of carbonmonoxide .57W. Traube and W. Lmge (and, in part, R.Stahn, R. Justh, and P.Baumgarten), BeT., 1925,58, [B], 2773; A,, 1926,257; compare Ann. Repwts,1912, 9, 44.s 7 W. A. Bone and F. R. Weston, Proc. Roy. SOC., 1926, [ A ] , 110, 615;A,, 480; W. A. Bone, R. P. Fraser, and D. M. Newitt, ibid., p. 634; A., 480INORGAMU CHEMISTRY. 63It has been found that in the chlorination of ferro-silicon the yieldof silicon hexachloride may be increased somewhat by adding silicontetrachloride to the chlorine ; this is, of course, in accordance withMartin's conclusions as to the mechanism of this chIorinati~n.~~By gradually adding to liquid ammonia a solution of disiliconhexachloride in anhydrous ether, ammonium chloride and diaminodi-iminodisilane, NH,.Si(:NH).Si(:NH).NH,, are formed. At -10")this compound loses ammonia to form polymeric tri-iminodisilane,Sf(iNH)>NH, which is stable at the ordinary temperature but de-Si(.NHIcomposes above 400" with some formation of silicocyanogen, Si,N,.These compounds are extremely sensitive to oxygen and moisture.From the products of the action of magnesium phenyl bromide ondisilicon hexachloride, dichlorodiphenylmonosilane, SiPh,Cl,, b.p.166"/17 mm., has been is0lated.~9In the course of further experiments upon silicic acids, the curiousobservation has been made that a certain form of silicic acid isvolatile in steam.60Pure germanochloroform has been prepared by the action ofhydrogen chloride on germanium dichloride; it has m. p. -71")b. p. 75*2", dw 1.93, and the vapour pressure has been measuredover the range -25" to 78.3".Decomposition of the compoundbegins a t 140" and is rapid a t 170") a t first by dissociation to yieldGeCI, + HC1, later to give the tetrachloride and metallic germanium.It is oxidised, even a t 0") by oxygen, probably according to thescheme : 4GeHC1, + 0, = 2GeC1, + 2GeC1, + 2H20.61 German-ium tetrachloride when pure, or in ethereal solution, reacts with dryammonia to give hexamminogermanic chloride, [Ge,GNH,]Cl,, as awhite powder which has no appreciable ammonia pressure a t theordinary temperature and may be kept for some days over con-centrated sulphuric acid without loss of weight. In aqueoussolution or in moist air it is slowly hydrolysed forming germanichydroxide. When the solid is treated a t 0' with ammonia at 3 atm.,it forms a second ammine, GeCl,,lGNH,, as a colourless liquidhaving an ammonia pressure of 760 mm. a t -4".Compoundsanalogous to the hexammine have been obtained with mono-, di-,and tri-ethylamines, propylamine, and butylamine.62Tertiary stannous phosphate, Sn,(PO,),, is a white, amorphousJ. I3. Quig and J. A. Wilkinson, J . Amer. Citem. Soc., 1926, 48, 902;A,, 589; compare G. Martin, J., 1914, 105, 2836.68 R. Schwarz and W. Sexauer, Ber., 1926, 59, [B], 333; A,, 369.6o R. Willstatter, H. Kraub, and K. Lobinger, ibid., 1925, 58, [B], 2462;L. M. Dennis, W. R. Orndorff, and D. L. Tabern, J. Pkysical Chem.,A., 1926, 36.1926, 30, 1049; A., 924.Oa W. Pugh and J. S. Thomas, J., 1926, 1051; A., 69564 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.powder, 4Z4* 3.823, insoluble in water, but soluble in mineralacids and alkali hydroxides ; it is produced by adding a 10% solu-tion of disodium hydrogen phosphate to a cold lo?/, solution ofstannous sulphate containing a little sulphuric acid.Stannoushydrogen phosphate, SnHPO,, crystallises in colourless tablets,d::''" 3476, from the solution obtained by dissolving granulatedtin in phosphoric acid (d 1.23), or in small, silky crystals on addingwater to a solution obtained by dissolving tin in phosphoric acid(d 1.5). Stannous dihydrogen phosphate, Sn(H,PO,),, results onheating the previous salt with phosphoric acid a t 140" and coolingthe solution over phosphoric oxide ; it crystallises in the form ofhighly refractive rhombs, dF'80 3.167, which are readily decom-posed by water.Stannous pyrophosphate, Sn,P,O,, is obtained asa white powder, 4.009, when the monohydrogen phosphate isheated a t 380-400" in a current of carbon dioxide. Stannousmetaphosphate, Sn(PO,),, is a white, glassy mass, d:Fg 3038, formedby heating the dihydrogen phosphate a t 390" in a current of carbondioxide. The stannous phosphates are more readily hydrolyseclthan the corresponding lead compounds, but otherwise are relativelystable.63Pure zirconium metal has becn obtained by preparing the puretetraiodide, resubliming it in a closed apparatus, and volatilisingit a t 600" in a vessel containing a fine tungsten wire heated electric-ally to 1800". Under these conditions, the iodide dissociates andpure zirconium is deposited as a rod, of which the tungsten wireforms only O.Olyo by volume.It has m. p. 2200" Abs., d 6.5, iscomparable with copper in ductility, and retains its lustre in air a tthe ordinary temperature but is superficially oxidised a t hightemperatures.64A new method of separating hafnium from zirconium makesuse of the fact that freshly-precipitated zirconium phosphate issoluble in oxalic acid and, on adding hydrochloric or sulphuricacid, is reprecipitated in a form which is readily filtered and affords acomparatively rapid separation of the hafnium in the precipitates.100 Kg. of a preparation containing 476% of zirconium and less than0.5% of hafnium were treated, and after 26 fractionations hafniumcontaining not more than 1% of zirconium was obtained.Thehafnium content of each fraction was determined with an X-rayspectrograph and confirmed chemically at intervals. This methodof separation has the advantages over fractional crystallisation ofthe double fluorides that it is more rapid, and that the hafniumA., 585.Is K. Jablczynslii and IY. Wiyckovslii, 2. anorg. Chem., 1926, 152, 207;J. L1. de Boer and J. D. Fast, ibid., 158, 1; A,, 699INORGANIC CHEMISTRY. 65accumulates in the less instead of in the more soluble fraction.From the pure hafnium phosphate, the metal (d 12.1) was preparedvia the hydroxide, oxide, chloride, metal, iodide, metal.65Some most interesting work is reported by Smits. A quartzapparatus, resembling a mercury-arc lamp but heated externally bygas-burners, was filled with molten lead of the highest purity andrun as a lead-arc lamp, in which, by rocking the tube, the arc wasmade and broken several times per second.At first, the lampshowed the lead spectrum only, but after running a t 80 volts and40 amp. for 10 hours, the spectrum of the light emitted showedstrong lines of mercury and thallium, and in experiments in whichsparking was employed to obtain high current densities, spectrashowing all the principal lines of mercury were obtained. In otherexperiments, a heavy spark-discharge between lead electrodesimmersed in carbon disulphide produced a fine deposit of dispersedlead. This lead was collected and heated in air and the distillate,when treated with iodine vapour, gave visible traces of mercuriciodide ; the same test applied to the lead electrode material gavenegative results.The earlier experiments with the quartz-lead lamp were neces-sarily of short duration, as the tubes were so blackened by a filmof lead silicate and silicon that observation became impossibleafter a few hours.A modified design of lamp now employed permitsmuch longer runs and the light after 39 hours shows a very strongmercury spectrum. In yet another experiment, 850 g. of leadremoved from a lamp after being used to produce an intermittentarc for 188 hours were heated to 800" in a quartz apparatus in acurrent of pure nitrogen, which then passed through two U-tubescooled in liquid air where 5 mg. of mercury condensed.Preciselysimilar treatment of the same quantity of the same sample of originallead which had not been used in a lamp gave no trace of mercury.These results are interpreted to indicate that a transmutation oflead into thallium and mercury has occurred, and whilst, of course,their general acceptance must await independent confirmation, theexperiments here referred to do seem to afford strong evidence oftransmutation.66Group V.The heat of formation of active nitrogen has been determined bycausing a current of activated gas to react with nitric oxide in acalorimeter and measuring the heat evolved and the amount of6 5 J. H. de Boer, 2. anorg. Chena., 1926,150, 210; A,, 373.66 A. Smits end A. Karssen, 2. Elelctrochem., 1926, 32, 578; A. Smita,Nature, 1926, 117, 13; A,, 106; compare idem, 2.anorg. Chew., 1926, 155,369; A,, 1016; (Miss) .4. C. Davies and F. Horton, Nature, 1926, 117, 152;A., 221; A. Smite., ibid., p. 620; A., 554.REF.-VOL. XXIII. 66 m& REPORTS ON THE PROQREBS OF CHEMISTRY.nitrogen peroxide produced. The mean value obtained for the heatof formation, 42,500 cal./g.-mol. (about 2-0 volts), supports thehypothesis that active nitrogen is nitrogen in a metastable molecularform. Observations of the effect of various gases in extinguishingthe fluorescence of active nitrogen indicate that only those for whichthe critical increments are less than a value approximately equal tothe above heat of formation have a positive effect.67 On the otherhand, the second positive group of bands, which predominates inthe discharge on activation of nitrogen, is believed to be due to theatom.68Nitrous oxide has been synthesised by passing an electric dis-charge through nitrogen a t a low pressure contained in a tube offused silica, the walls of which had previously been saturated withoxygen by passing a discharge through the tube lilled with thatgas.The nitrous oxide was isolated as i t was formed by condensingit in a U-tube surrounded with liquid air.60In the course of measurements of its compressibility up to 160atm. and over the temperature range -SO" to lo", it has been ob-served that under prolonged compression nitric oxide decomposes,yielding a blue, liquid mixture of nitrous oxide and nitrous anhydride.At 700 atm.the decomposition is very rapid. Freshly-preparedand liquefied nitric oxide is but faintly blue in colour, and the puresubstance would probably be colourless, but multiple liquefactiondoes not effect purification, since the intensity of the blue colouris thereby gradually increased.50 On the other hand, nitric oxideis formed in appreciable amounts (up to 25% of that required by theequation 2N,O + 2NO + N,) by dissociation of nitrous oxideat 1300". At lower temperatures (700"), the decomposition isslower and proceeds chiefly according to the exothermic reaction,2N,O + 2N, + 0, ; the former, slightly endothermic, reaction isfavoured by rise of ternperat~re.~~It has been found that the decomposition of nitrogen pentoxideis not accelerated by infra-red radiation corresponding with its strongabsorption bands ; this is contrary to the predictions of the radiationtheory.72 A study of the thermal decomposition of nitrogen pent-6 7 E.J. B. Willey and E. K. Rideal, J., 1926, 1804; A., 893; E. J. B.Willey, Nature, 1926, 117, 381; A., 336; ibid., 118, 735; A., 1213.M. Duffieux, ibid., 117, 302; A., 336; compare (Lord) Rayleigh, ibid.,p. 381; A., 336.D. L. Chapman, R. A. Goodman, and R. T. Shepherd, J., 1936, 1404;E. Briner, H. Biedermann, and A. Rothen, He2v. Chiin. Acta, 1935, 8,Aq 811.923; A., 1926, 16.'l E. Briner, C. Meiner, and A. Rothen, ibid., 1926, 9, 409; A., 686.72 H. A. Taylor, J . Amer. Chem. Soc., 1926, 48, 677; A., 485; compareF. Daniels, ibid., p. 607; A., 485INORGANIC CHEMISTRY.67oxide a t low pressures has shown that dissociation is not retardedby lowering of pressure, but below a critical pressure, approxim-ately 0.25 mm., it is gradually accelerated with falling pressure until,a t the lowest pressures, the rate of dissociation becomes approxim-ately constant a t about five times the normal value. This curiousobservation is quite irreconcilable with all '' chain " mechanisms,and is most readily interpreted on the assumption that a definitefraction of the activated molecules always undergoes decompositionirrespective of pressure, but that a larger fraction (about four-flfths)does not decompose if i t collides within 10-6 see. after activation,but is deactivated by collision. Calculations on this basis are ingood agreement with experiment.A possible explanation of thetwo different types of activated molecules is offered, based on thefact that there are four N-0 linkings, and only one shared N-0linking in the molecule of the pentoxide. Activation of one linkingcauses decomposition-activation of the shared linking invariablyso; activation of the others, only after a time interval and ifcollisions do not intervene.73A number of reactions of elements electronegative to nickelupon ammonio-bases in liquid ammonia solution have been investig-ated : the initial reaction is of the type C1, + 2KOH j KC1 +KClO + H,O. By the action of sodamide or potassamicle uponexcess of tin amalgam, new sodium and potassium ammoniostann-ites, Na (or K)[Sn(NH,)J, have been obtained.'* A break a t 310"in the decomposition curve of the compound PCl,,lONH, corre-sponds with a break observed a t the same temperature with amixture of ammonium chloride and sand, and thus suggests thatthe substance is partly a mixture of ammonium chloride and theamine P(NH,),-a view which is supported by the action of liquidammonia on the sul~stance.~~Pyridine hydrazinedisulphonate is prepared in S0-85% yieldby the action of chlorosulphonic acid on a, suspension of hydrazinesulphate in cold pyridine and subsequent precipitation of the saltby addition of ethyl alcohol ; the corresponding ammonium (+ H,O)and sodium (+ 2H,O) salts are described.The pyridine salt isoxidised by sodium hypochlorite in the presence of water a t -20"to the azo-compound, isolated by addition of potassium chloride aspotassium azodisulphonate, S0,K.N:N.S03K.70Red or yellow phosphorus reacts with water a t 238-360" and'3 H.S. Hirst and E. K. Rideal, Proc. Roy. Soc., 1925, [ A ] , 109, 626; A.,74 F. W. Bergstrom, 6. Physical Chem., 1926, 30, 12; A., 264.7 6 H. PerpBrot, Bull. SOC. ckim., 1925, [iv], '37, 1640; A., 1926, 137.76 E. Konrad and L. Pellens, Ber., 1926, 69, [B], 136; A., 370.1926, 3268 ANNUAL REPORTS ON THE PROQRESS OF CHEMISTRY.57-360 atm. to form phosphine and orthophosphoric acid thus :4P, + 12H20 + 3H,PO, + 5PH,. If hydrogen be added, theproportion of phosphine increases until, with dry hydrogen at 360"or less, it is the sole product. When phosphorus is heated withwater for a short time a t 248" and 48 atm.and the reaction is rapidlyinterrupted, there is formed a crystalline purple phosphorus, d 1.93,which has an ignition temperature 210". At higher temperatures,and pressures (360-380" and )90 atm.) crystalline black phos-phorus, d 3.06, is formed, apparently by decomposition of phosphinefirst formed.The velocity of air-blast required to remove the glow from phos-phorus and maintain it downstream may be used as a measure ofthe rate of propagation of the glow. Using this method in an exten-sion of Rayleigh's experiments, it has been found that inhibitors,such as ethylene, benzene, chloroform, and aniline, diminish the rateof propagation, and to an extent which is lessened by rise oftemperature, and thus act in this matter just as oxygen d0es.7~The addition of stannic chloride, either as solid or in very con-centrated solution, to a concentrated solution of sodium hypophos-phite produces a voluminous, white precipitate of a strongly reduc-ing substance, the analysis of which corresponds with the formulaSnCI,,Sn( H2P0,),,3H,0.When heated, the substance is dehydrateda t 140", and decomposes a t 190" with a characteristic reddeningand the evolution of phosphine. If the heating is stopped imme-diately the red colour appears, and the mass extracted with con-centrated hydrochloric acid, a bright red residue remains, whichhas very strong reducing properties and a composition approximat-ing closely to that of phosphorus suboxide, P40.79A large number of new complex metallic phosphites and pyro-phosphates have been prepared.soPhosphorus nitride, PN, has been obtained by passing a dischargebetween aluminium electrodes in nitrogen a t about 200 mm.in atube lined with yellow phosphorus, extracting the product withcarbon disulphide, and heating the residue in porcelain a t 550"in a stream of nitrogen a t 12 mm. pressure. The residual nitride isa voluminous, yellowish-brown powder, very resistant to chemicalagents .81The preparation of an acid barium vanadate, 2Ba0,3V20,,12H,O,7 7 V. Ipatiev and W. Nikolajev, Ber., 1926, 59, [B], 695; A,, 487.H. J. Emelhs, J., 1926, 1336; A., 777.A. Terni and C. Padovani. Atti R. Accad. Lincei, 1925, [vi], 2, 501; A.,A. Rosenheim, S.Frommer, H. Gliiser, and W. Hiindler, 2. anorg. Chem.,1926, 265.1926, 153, 126; A., 696.81 W. Moldenhauer and H. Dorsam, Ber., 1926, 59, [B], 926; A., 696INORGANIC CHEMISTRY. 69has been described ; and from this salt, by interaction with metallicsalts, similar vanadates of nickel, cobalt, copper, beryllium, cad-mium, and manganese are obtained. They take up gaseous ammoniaslowly, and combine more rapidly with liquid ammonia, to formhexammines.82Group V I .Excess of hydrogen sulphide acts upon 1% aqueous potassiumpermanganate according to the scheme 10KMn0, + 22H,S +3K,S04 + lOMnS + 2K,S,03 + 22H20 + 5s ; in the earlier stagesof the reaction some dithionate is formed.83 Experiments upon thedecomposition of dilute aqueous thiosulphuric acid have shown thatthe yellow colour characteristic of such solutions is due to a sulphurcompound derived from the oxide S,03; about 407; of the totalsulphur present exists in this formas4 In contirmation of this, it isfound that anhydrous thiosulphates with liquid sulphur dioxide a tlow temperatures yield yellow solid compounds, K,S,O,,SO, andRb,S,O,,SO,, which give clear yellow aqueous solutions in whichthe equilibrium S,O," + SO,The method of intensive drying has been applied to a furtherstudy of the complexity of the solid state; repeated distillationof the dried ice-like form of sulphur trioxide yields the high-melting(as distinct from the low-melting) asbestos-like form, and this givesmarked evidence of complexity by a great decrease of vapourpressure (from 591 to 37 mm.) on distillation.This low vapourpressure is substantially constant a t 18", but a t higher temperaturesit increases, approaching asymptotically the value corresponding toinner equilibrium. The establishment of equilibrium is hastenedby incidence of X-rays, and all forms yield identical X-ray photo-graphs. The original papers should be consulted for a full accountof the experimental method and of the argument directed toestablish the great complexity of the system formed by this simplecompound.86It is convenient here to mention other work of a kindred nature.To determine whether intensive drying results in a fixing of theinner equilibrium, or a displacement of the inner equilibriumfollowed by fixation, measurements of the vapour pressure ofintensively dried nitrogen tetroxide and m-hexane have been made.Nitrogen tetroxide, after intensive drying for 23 months a t the(S,O,(SO,)]" exists.858 3 F.Ephraim and G. Beck, Helv. Chim. Acta, 1926, 9, 35; A., 370.83 H. B. Duaaicliff and S. D. Nijhawan, J., 1926, 1; A., 256.84 E. H. Riesenfeld and E. Grimthal, Medd. K. Vetenskapsukud. NobeLInst.,O 5 F. Foerster and R. Vogel, Z . anorg. Chern., 1926, 155, 161; A., 1016.8 8 A. Smits and P. Schoenmaker, J . , 1926, 1108, 1603; A., 669, 785.1925, 6, No. 9, 1 ; A., 1926, 25770 ANNUAL REPORTS ON THE PROQRESS OF CHEMISTRY.ordinary temperature in an apparatus the glass of which was notfreed from capillaries, showed an increase in the vapour pressure of1.9 cm.of mercury; a part of the liquid was then distilled off, andthe increase fell to 0-4 cm., but after a further 11 days this hadrisen to 1.47 cm. In another apparatus made of capillary-free glass,a much more rapid drying effect was obtained, for after 16 monthsthe vapour pressure had risen by 3.3 cm. On raising the temper-ature, the increase passed through a maximum, as had been antici-pated on theoretical grounds. The changes of vapour pressure, andof colour to a deeper brownish-red, prove that the drying processhad effected a displacement of the inner equilibrium in the directionN,04 + 2N0,.With n-hexane, intensive drying a t about 40" results in a decreaseof vapour pressure. After only 14 weeks, this reached 0.9 om.,showing that, in contrast to nitrogen tetroxide, the iilner equilibriumis here shifted towards the less volatile component.The decreasewas augmented by raising the temperature, and after distilling offa portion of the liquid a t the drying temperature, the vapourpressure tended to revert to its previous value, showing that theinner equilibrium had not been reached.s7Pyrosulphuryl chloride has been prepared by the action of carbontetrachloride on sulphuric or chlorosulphonic acid a t SO", and is ahygroscopic substance, 1.834, n:"" 1,449, b. p. 57'130 mm.,52'115 mm., having a characteristic odour. When heated, it dis-sociates irreversibly according to the schemes S,O,Cl, --+ SO, +SO, + C1, and S,O,Cl, -+ SO, + SO,Cl,. Measurements of itsdiamagnetism and molecular refraction indicate the constitution :When chromium trioxide is heated a t 263") oxygen is evolved andthe liquid slowly solidifies to a dark brownish-violet mass of a newoxide CrsO1,, which, like the known oxide, Cr5012, appears to be achromate of chromium.89 The chromium carbonyl, Cr(CO),, hasbeen prepared by slowly adding ethereal magnesium phenyl bromideto an ether-benzene suspension of chromium chloride in presence ofcarbon monoxide belowSelenic-uranic acid, H,[U03(Se04)],2H,0, diselenic-uranic acid,H,[UO4(SeO4),],2H,O (and with 6H,O), and triselenic-diuranic acid,8 7 A.Smits, W. de Liefde, E. Swart, and A. Claassen, J . , 1926,2657; A.,1206; A. Smits, ibid., p. 2655; A., 1206; compare S. B. Mali, 2. anorg.Chem., 1925, 149, 150; A,, 1926, 117; J.W. Williams, J. Arne?. Chem. Soc.,1925, 47, 2644; A,, 1926, 15.88 V. Grignard and P. Muret, Compt. rend., 1926, 183, 581; A,, 1113;ibid., p. 713; A., 1218.s(:o)(ocl)*o~s(:o)(ocl).~~A. Simon and T. Schmidt, 2. anwg. Chem., 1926, 153, 191; A., 697.go A. Job and A. Cased, Compt. rend., 1926, 183, 392; A., 1017INORGANIC CHEMIETRY. 71H6[U02(U04)(Se04)3],7H,0, and a number of potassium andammonium salts derived therefrom, have been prepared.91By heating the known uranium nitride, U3N4, a t 1740" and 1900",two new nitrides, U,N4 and U,N,, are obtained : these differ fromU,N, in being completely soluble in acids with the evolution ofhydrogen and conversion of the nitrogen into ammonium salts.92Group V I I .Several oxidising reactions of fluorine have been investigated.Fluorine with aqueous ammonium hydrogen sulphate or drypotassium hydrogen sulphate yields considerable amounts of thepersulphates, in the latter case accompanied by potassium fluoro.sulphonate, KFS03.gSHydrogenchloride is found to combine readily with certain metallicsulphates, forming compounds of the type XS04,2HC1 ; this occurschiefly in the cases where the metallic chlorides do not readily yieldhydrogen chloride when treated with concentrated sulphuric acid,and the temperature a t which the evolution of gas begins is approxim-ately the dissociation temperature of the complex.94 Determinationshave been made of the vapour pressure of pure chlorine dioxide fromthe f .p. -59" t o the b.p. +l1°.95In an interesting investigation of the influence of water on theunion of halogens with hydrogen, it has been observed that whilstthe combination of hydrogen and iodine can be inhibited by inten-sive drying, the decomposition of hydrogen iodide, similarly driedand a t the same temperature, proceeds apparently unhindered.Hence, in this case, as in those previously mentioned, intensive dry-ing skifts the position of equilibriumag6Iodic acid and stannic chloride in dilute nitric acid solution yield,according to the conditions, dihydroxytetraiodatostannic acid,Sn(IO,),(OH),H,, or stanni-iodic acid, Sn(I03),H2. The alkali-metal salts of the latter acid and also of tetrahydroxydi-iodato-stannic acid, Sn(103)2(0H)4H2, have been prepared.Iodic acid andantimony pentachloride yield trihydroxytri-iodatoantimonic acid,Sb(I03)3(OH)3H.9791 J. Meyer and E. Kasper, 2. anorg. Chem., 1926, 155, 49; A., 926.92 0. Heuder, ibid., 154, 363; A., 909.98 F. Fichter and K. Humpert, Helv. Chim. Acta, 1926, 0, 467, 602, 692;a* F. Ephraim, Ber., 1925, 58, [B], 2262; A,, 1926, 36; itid., 59, [B], 790;95 F. E. King and J. R. Partington, J., 1926, 625; A.. 569.98 B. Lewis and E. I<. Rideal, J . Anzev. Chena. Soc., 1926, 48, 2553; A,,$ 7 P. R$y and S. N. Ray, J. Indian Chem. Soc., 1926, 3, 110; A,, 1015.A., 699, 925.A,, 587.111172 ANNUAL REPORTS ON THE PROQRESS OF CHEMISTRY.Although a good deal of discussion has taken place upon thesupposed discovery of dvi-manganese previously reported, nothingsubstantial has been added to our knowledge of the element.gBGroup V I I I .New hydrides of cobalt, iron, and chromium have been obtainedby the interaction of the chlorides of those metals with hydrogenin presence of magnesium phenyl bromide.Cobalt hydride has theformula CoH,; iron dihydride, FeH,, is a black powder; ironhexahydride, FeH,, is a black, viscous oil; chromium trihydride,CrH,, is a black p r e ~ i p i t a t e . ~ ~A study of the composition of various ferro- and ferri-cyanides,prepared in a variety of ways, has led to the conclusion that ordinaryPrussian blue is the ferrous salt of a complex acid formulated asThe identity of the absorption spectra of ferronitricoxide salts inaqueous solution confirms the view that they all give the samecoloured cation, FeNO' : and the similarity between these spectraand those of the black series of Roussin's salts lends support to theview that the latter should be formulated 2A number of interesting papers on complex cobalt salts cannotbe summarieed in the space available and must be dismissed with areference ?Platinum tribromide is obtained by heating platinum tetrabromidebetween 370" and 405" ; above 410°, the tribromide loses bromineto form the dibromide, but this is stable over a range of only 5'.J.Heyrovsk9, Nature, 1926, 117, 16; A,, 138; J. G. F. Druce, ibid.,p. 16; A., 138; V. Dolejitek, J. G. F. Druoe, and J. Heyrovsky, ibid., p. 159;A., 227; V. Dolejgek and J. Heyrovskf, Chem. Listy, 1926, 20, 4 ; A,, 258;F.H. Loring, Chem. News, 1926, 152, 101; A., 338; 0. Zvjaginstsev, M.Korsunski, and N. Seljakov, Nature, 1926, 118, 262; A., 934; B. Polland,Con~pt. rend., 1926, 183, 737; A., 1194; F. H. Loring and J. G. F. Druce,Chem. News, 1925, 131, 337; A., 12.Q Q T. Weiohselfelder and €3. Thiede, Annulen, 1926, 447, 64; A., 372.N. Tamgi, Cfazzetta, 1925, 55, 951; A., 1926, 259.W. Manchot and E. Linckh, Ber., 1926, 59, [B], 406, 412; A., 452, 453.E. H. Riesenfeld (with W. Petrich), iMedd. K . Vetenskapsakad. Nobel-Inst., 1925, 6, No. 6, 1 ; A., 1926, 259; R. Klement, 2. anorg. Chem., 1926,150, 117; A., 372; F. L. Hahn, H. A. Meier, and H. Siegert, ibid., p. 126;A., 372 ; F. M. Jaeger and P. Koets, Proc. K. Akad. Wetensch. Amsterdam, 1926,29, 5 9 ; A., 697; J. Meyer and K. Grohler, 2. anorg. Chem., 1926, 155, 91;A., 925; B. K. Paul and P. V. Sarkar, A m . Chim., 1926, [XI, 5, 199; A., 588INORGANIC CHEJIISTRY. 73Platinum tetraiodide is produced by heating platinum with iodinein a sealed tube a t 240-300", and this is converted to the tri-iodidea t 350400°.4Tensimetric determinations of the molecular weights of cis- andtrans-dichlorodiamminoplatinum in liquid ammonia show that thecis-salt has a normal and the trans-salt a double molecular weight.The latter salt is therefore formulated :The action of hydrogen peroxide or ozone on complex compoundsof bivalent platinum renders the platinum quadrivalent and therebyadds two negative groups to it. Thus, from Peyronne's chloride areobtained the dihydroxy-compound, [Pt,SNH,,Cl,(OH),], and a newhydroxypentamminoplatinic carbonate, [Pt,5NH,,0H],(C03),.gBy passing carbon monoxide into a suspension of palladouschloride in anhydrous alcohol at 0", or by passing the gas, chargedwith methyl alcohol vapour, over palladous chloride a t 15", the com-pound PdCl,,CO is obtained. It readily decomposes into palladiumand carbon dioxide.7Ruthenium tetrachloride, RuC1,,5H20, has been obtained ashygroscopic, reddish-brown, monoclinic crystals. It is formed whena solution of hydroxytetrachlororuthenic acid in concentratedhydrochloric acid is saturated with chlorine for 5 hours a t 100" andthen evaporated in a current of dry chlorine. It is hydrolysed indilute solution, but in a strong solution is stable a t 15", does notliberate iodine from potassium iodide, gives a black precipitate withsilver nitrate and forms crystalline double salts with potassium andammonium chlorides.*Two new chlorides of rhodium, RhCl and Rha,, result from theaction of chlorine on rhodium a t 948-96gb; and by dissociationof rhodium trioxide a t 750" in oxygen, two new oxides, Rh,O andRho, insoluble in acids, but readily reducible by hydrogen, havebeen ~ b t a i n e d . ~H. V. A. BRISOOE.L. Wohler and F. Miiller, 2. anwg. Chem., 1925, 149, 377; A., 1926, 259.H. Reihlen and K. T. Nestle, Annalen, 1926, 447, 211; A., 699.L. Tschugaev and W. Chlopin (with E. Fritzmann), 2. anwg. C?bem.,' W. Manchot and J. Konig, Ber., 1926, 50, [B], 883; A,, 698; compare1926, 151, 253; A., 373.W. Manchot, ibid., 1925, 58, [B], 4518; A., 1926, 138.S. Aoyama, 2. anorg. Chem., 1926, 153, 246; A., 698.L. Wohler and W. Muller, ibid., 1925, 149, 125; A,, 1926, 138.C